Abstract
Social news sites where users actively engage in reading, discussing, and sharing news with their network can serve as a rich dataset for observing and analyzing the behavior of online social news consumption. In this paper, we combine machine learning and network analysis of users textual contents and network characteristics to propose metric that measures users degree of seeking diversity in a social new site. Our results reveal that the proposed metric serve to identify influential users who span structural holes and promote to create smaller information network. We discuss this result using a dataset of Huffington Post articles from the Politics section containing over 43,000 articles and activities of over 35,000 users.
| Original language | English |
|---|---|
| Title of host publication | Experiments with Non-parametric Topic Models |
| Place of Publication | USA |
| Publisher | Association for Computing Machinery (ACM) |
| Edition | Peer Reviewed |
| ISBN (Print) | 9781450329569 |
| Publication status | Published - 2014 |
| Event | 20th ACM SIGKDD Conference on Knowledge Discovery and Data Mining KDD2014 - New York, USA, United States Duration: 1 Jan 2014 → … http://www.kdd.org/kdd2014/ |
Conference
| Conference | 20th ACM SIGKDD Conference on Knowledge Discovery and Data Mining KDD2014 |
|---|---|
| Country/Territory | United States |
| Period | 1/01/14 → … |
| Other | August 24-27 2014 |
| Internet address |